Each OMO agent gets its own model
Route each OMO agent to a different model. Guide to Oh My OpenCode's default agent→model routing.
what to steal
Route by role: put the expensive model only where reasoning matters.
Our takeaway from the source, not a quote.
rhcwlq89-omo-routing.toml
# Route each OMO agent to a different model # source: https://rhcwlq89.github.io/en/blog/oh-my-opencode-guide/ [rig] who = "Guide to Oh My OpenCode's default agent→model routing" author = "rhcwlq89" date = "2026-02-08" harness = ["OpenCode", "OMO (Oh My OpenAgent)"] models = ["Claude Opus 4.5 High (Sisyphus)", "GPT 5.2 Codex (Hephaestus)", "GPT 5.2 (Oracle)", "Gemini 3 Pro (Frontend)", "Claude Sonnet 4.5 (Librarian)", "Claude Haiku 4.5 (Explore)"] [keep] - Per-agent model routing
⚠ Claude Opus 4.5 High, GPT 5.2 Codex, GPT 5.2, Gemini 3 Pro, Claude Sonnet 4.5, Claude Haiku 4.5: older model(s), named by the source on 2026-02-08. See current models.
Source: https://rhcwlq89.github.io/en/blog/oh-my-opencode-guide/ · checked 2026-10-14
// similar rigs
~/rigs/veladan-opencode-local.toml2026-06-01
OpenCode + OMO, 100% local
steal → Pair a local agent with a self-hosted forge like Gitea and nothing leaves the box.
OpenCodeOMO (Oh My OpenAgent) opencode/minimax-m3-fr
by Veladan
~/rigs/michyamrane-omo-slim-free.tomlundated
A full OpenCode agent team on free models
steal → Give every agent a fallback array so a rate limit never stops the run.
OpenCodeOMO (Oh My OpenAgent) Free models
by mIcHyAmRaNe
~/rigs/joekarlsson-opencode-litellm.toml2026-10
Broke up with Claude Code: open-weight models at about two cents a task
steal → Put every model call behind one gateway with per-app budgets, and force a test run before 'done'.
OpenCodeClaude Code DeepSeek V4.1 FlashClaude Claude Code as default
by Joe Karlsson